Pith. sign in

REVIEW 2 major objections 1 minor 78 references

Order Matters: Unveiling the Hidden Impact of Macro Placement Sequences via Proxy-Guided LLM Evolution

T0 review · 2 major / 1 minor · reviewed 2026-06-27 · grok-4.3

Pith's one-line read The sequence in which macros are placed during chip design is a decisive optimization factor that LLM evolution can automatically improve.

desk verdict OrderPlace's claimed gains rest on an unverified proxy whose ranking fidelity is not demonstrated, so the results stay provisional. read the letter →

arxiv 2606.08904 v1 pith:BSC455Z4 submitted 2026-06-08 cs.AI

classification cs.AI
keywords macroplacementsequencingLLMevolutionproxyevaluationphysicaldesignwirelengthoptimizationcombinatorial
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper shows that placement order is not a minor preprocessing choice but a major driver of final solution quality, because early decisions create irreversible constraints on later placements. OrderPlace evolves new ordering policies with large language models, ranging from static metrics to dynamic mechanisms, and uses a fast deterministic greedy probe as a proxy to screen candidates before full evaluation. On the ISPD 2005 benchmarks this yields sequences that cut wirelength by 34.04 percent versus WireMask-EA and 14.08 percent versus EGPlace. A reader would care because the work treats sequencing as an optimizable dimension rather than a fixed heuristic, opening a route to better layouts without rewriting the core placer.

What carries the argument

Proxy-guided LLM evolution that generates and screens placement-ordering policies via a deterministic greedy probe used as a lightweight filter.

What would settle it

A sequence that the proxy ranks low but that produces lower wirelength than the proxy's top-ranked sequences when evaluated with the full placer would falsify the filtering claim.

Watch

Extended reading notes

Core claim

OrderPlace demonstrates that placement sequences are a decisive factor in macro placement optimization; proxy-guided LLM evolution automatically discovers code-level ordering policies that produce lower wirelength than prior methods, with measured reductions of 34.04 percent relative to WireMask-EA and 14.08 percent relative to EGPlace on standard ISPD 2005 benchmarks.

Load-bearing premise

The greedy probe proxy ranks sequences accurately enough that the best full-evaluation performers are not discarded during filtering.

Editorial extensions

If this is right

  • Placement ordering can be treated as an explicit search dimension rather than a static heuristic based on area or connectivity.
  • Early sequencing choices create domino effects that permanently limit the reachable solution space.
  • Lightweight proxies make evolutionary discovery of ordering policies computationally practical.
  • Novel dynamic and physics-inspired ordering rules can be found automatically instead of being hand-crafted.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same proxy-guided evolution approach could be tested on sequencing decisions inside other stages of physical design such as routing or floorplanning.
  • Automatically discovered ordering policies might transfer across different benchmark suites or technology nodes without manual retuning.
  • Combining evolved ordering with existing ML coordinate predictors could produce compound gains in overall placement quality.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces OrderPlace, a proxy-guided LLM evolution framework for automatically discovering macro placement order strategies in chip physical design. It claims that placement sequencing is a decisive factor due to irreversible early decisions, explores code-level policies via LLM evolution, and uses a deterministic greedy probe as a lightweight proxy to filter candidates. On ISPD 2005 benchmarks, OrderPlace is reported to reduce wirelength by 34.04% compared to WireMask-EA and 14.08% compared to EGPlace, attributing the gains to novel discovered ordering strategies.

Significance. If the proxy mechanism is shown to produce rankings correlated with full placement evaluations, the work would be significant for electronic design automation by demonstrating that temporal ordering can be optimized beyond static heuristics and that LLM-based search can discover effective policies for high-dimensional combinatorial problems. The approach of evolving code-level mechanisms rather than fixed metrics offers a generalizable direction for heuristic discovery in VLSI optimization.

major comments (2)
  1. [Methods (proxy evaluation mechanism)] Methods section on proxy evaluation: the deterministic greedy probe proxy is presented without any reported quantitative validation (e.g., rank correlation, precision@K, or hold-out comparison against full wirelength evaluations on ISPD 2005 instances). This is load-bearing for the central claim because the evolutionary search filters all candidates exclusively via the proxy; any systematic mismatch would mean the reported novel strategies and wirelength gains are artifacts of the proxy rather than genuine improvements.
  2. [Experimental results] Experimental results section: the headline improvements (34.04% vs WireMask-EA, 14.08% vs EGPlace) are stated without error bars, number of independent runs, statistical significance tests, or details on how many sequences were evaluated under the full placer versus the proxy. This undermines assessment of whether the gains are robust or attributable to the discovered orderings.
minor comments (1)
  1. [Abstract] The abstract and methods do not specify the exact ISPD 2005 instances used or the full experimental protocol, which would aid reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on the proxy validation and experimental reporting. We address each major comment below and will revise the manuscript accordingly to strengthen the claims.

read point-by-point responses
  1. Referee: Methods section on proxy evaluation mechanism: the deterministic greedy probe proxy is presented without any reported quantitative validation (e.g., rank correlation, precision@K, or hold-out comparison against full wirelength evaluations on ISPD 2005 instances). This is load-bearing for the central claim because the evolutionary search filters all candidates exclusively via the proxy; any systematic mismatch would mean the reported novel strategies and wirelength gains are artifacts of the proxy rather than genuine improvements.

    Authors: We agree that explicit quantitative validation of the proxy is essential to support its use as a filter. In the revised manuscript, we will add a dedicated subsection reporting Spearman's rank correlation and Kendall's tau between proxy scores and full wirelength evaluations on all ISPD 2005 instances, along with precision@5 and precision@10 metrics and a hold-out comparison. These additions will directly demonstrate the proxy's ranking fidelity. revision: yes

  2. Referee: Experimental results section: the headline improvements (34.04% vs WireMask-EA, 14.08% vs EGPlace) are stated without error bars, number of independent runs, statistical significance tests, or details on how many sequences were evaluated under the full placer versus the proxy. This undermines assessment of whether the gains are robust or attributable to the discovered orderings.

    Authors: We acknowledge that the current experimental presentation lacks statistical rigor. In the revision, we will report results from 5 independent runs with standard deviations shown as error bars, include p-values from paired t-tests against the baselines, and explicitly state the number of sequences evaluated under the full placer (versus those filtered by proxy only). This will clarify the robustness of the reported wirelength gains. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: empirical search results with independent proxy filter

full rationale

The paper presents an LLM evolutionary search over ordering policies, filtered by a deterministic greedy proxy before full placement evaluation. No equations, fitted parameters, or self-citations are described that would make the reported wirelength reductions equivalent to the inputs by construction. The proxy is an engineering choice whose correlation with true quality is an external assumption, not a definitional loop. The headline gains are benchmark outcomes, not renamings or self-referential predictions.

Assumptions & free parameters 0 free parameters · 0 assumptions · 0 invented entities

Abstract-only review; no free parameters, axioms, or invented physical entities are stated. The OrderPlace framework itself is a new procedural method rather than a postulated entity with independent evidence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Order Matters: Unveiling the Hidden Impact of Macro Placement Sequences via Proxy-Guided LLM Evolution." pith.science (2026). https://pith.science/paper/BSC455Z4

@misc{pith2026260608904,
  author       = {Pith},
  title        = {Pith review of: Order Matters: Unveiling the Hidden Impact of Macro Placement Sequences via Proxy-Guided LLM Evolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BSC455Z4}},
  note         = {Machine review of arXiv:2606.08904}
}
read the original abstract

Macro placement is a fundamental step in modern chip physical design, playing a crucial role in determining the solution quality of high-dimensional combinatorial optimization problems. Despite recent advancements in machine learning for spatial coordinate determination, the temporal dimension of placement sequencing remains largely governed by static heuristics. In this work, we demonstrate that the placement sequence is not merely a preprocessing step but a decisive factor in optimization, where suboptimal early decisions trigger irreversible domino effects that constrain the solution space. To harness this unexplored dimension, we propose \textbf{OrderPlace}, a proxy-guided LLM evolution framework for automatically discovering macro placement order strategies. Instead of relying on manually crafted heuristics such as area- or connectivity-based ordering, OrderPlace explores a broader space of code-level policies, ranging from static scoring metrics to dynamic physics-inspired mechanisms. To mitigate the prohibitive cost of evaluating sequences, we introduce a lightweight proxy evaluation mechanism that efficiently filters candidates using a deterministic greedy probe. Experimental results on the standard ISPD 2005 benchmarks demonstrate that OrderPlace discovers novel ordering strategies. Compared with WireMask-EA and the state-of-the-art method EGPlace, OrderPlace reduces wirelength by 34.04\% and 14.08\%, respectively.

Figures

Figures reproduced from arXiv: 2606.08904 by the authors.

Figure 1
Figure 1. Overview of OrderPlace. Part (a) illustrates the overall execution flow of OrderPlace, while part (b) presents the detailed procedure of the Population Quality Evaluator module. (a) Select Position: Determine the optimal position pt = (xt, yt) for the current macro πt that mini￾mizes the incremental wirelength cost: pt = argmin p∈Valid(Pt−1) ∆HPWL(πt, p, B) (1) (b) Update Placement: Pt = Pt−1 ∪ {(πt, pt)}. (c) Updat… view at source ↗
Figure 2
Figure 2. Comparison of HPWL Trend Over the Runtime(s). In these trajectories, OrderPlace(StrategyNumber) denotes the placement process utilizing the Number-th best ordering strategy identified by our framework. For each dataset, we select the top 4 strategies and perform parallel macro placement optimization, resulting in a total of 24 strategies. For specific strategy details, refer to Appendix F. • Select the position mini… view at source ↗
Figure 3
Figure 3. Experimental Analysis Results for Parameter Top-K [PITH_FULL_IMAGE:figures/full_fig_p032_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Below subfigure is the convergence curve of the best proxy HPWL as the generation number increases, where the bar chart represents the elite population update rate for each generation. Above subfigure is the composition of the elite population for each generation, with…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

78 extracted references · 4 canonical work pages

  1. [1]

    Langley , title =

    P. Langley , title =. Proceedings of the 17th International Conference on Machine Learning (ICML 2000) , address =. 2000 , pages =

  2. [2]

    T. M. Mitchell. The Need for Biases in Learning Generalizations. 1980

  3. [3]

    M. J. Kearns , title =

  4. [4]

    Machine Learning: An Artificial Intelligence Approach, Vol. I. 1983

  5. [5]

    R. O. Duda and P. E. Hart and D. G. Stork. Pattern Classification. 2000

  6. [6]

    Suppressed for Anonymity , author=

  7. [7]

    Newell and P

    A. Newell and P. S. Rosenbloom. Mechanisms of Skill Acquisition and the Law of Practice. Cognitive Skills and Their Acquisition. 1981

  8. [8]

    A. L. Samuel. Some Studies in Machine Learning Using the Game of Checkers. IBM Journal of Research and Development. 1959

Show all 78 references
  1. [9]

    2000 , publisher=

    MacMillen, Don and Camposano, Raul and Hill, Dwight and Williams, Thomas W , journal=. 2000 , publisher=

  2. [10]

    Kim, Myung-Chul and Markov, Igor L , booktitle=

  3. [11]

    Wang, Laung-Terng and Chang, Yao-Wen and Cheng, Kwang-Ting Tim , year=

  4. [12]

    Shi, Yunqi and Xue, Ke and Lei, Song and Qian, Chao , journal=

  5. [13]

    Deng, Ji and Li, Zhao and Zhang, Ji and Gao, Jun , booktitle=

  6. [14]

    Lai, Yao and Mu, Yao and Luo, Ping , journal=

  7. [15]

    Geng, Zijie and Wang, Jie and Liu, Ziyan and Xu, Siyuan and Tang, Zhentao and Yuan, Mingxuan and Hao, Jianye and Zhang, Yongdong and Wu, Feng , booktitle=

  8. [16]

    Lai, Yao and Liu, Jinxin and Tang, Zhentao and Wang, Bin and Hao, Jianye and Luo, Ping , booktitle=

  9. [17]

    2021 , publisher=

    Mirhoseini, Azalia and Goldie, Anna and Yazgan, Mustafa and Jiang, Joe Wenjie and Songhori, Ebrahim and Wang, Shen and Lee, Young-Joon and Johnson, Eric and Pathak, Omkar and Nova, Azade and others , journal=. 2021 , publisher=

  10. [18]

    Lin, Yibo and Dhar, Shounak and Li, Wuxi and Ren, Haoxing and Khailany, Brucek and Pan, David Z , booktitle=

  11. [19]

    2024 , publisher=

    Majumdar, Somdeb and Mallappa, Uday and Mostafa, Hesham , journal=. 2024 , publisher=

  12. [20]

    Proceedings of the thirteenth International Conference on Artificial Intelligence and Statistics , pages=

    Ross, St. Proceedings of the thirteenth International Conference on Artificial Intelligence and Statistics , pages=

  13. [21]

    2011 , publisher=

    Kahng, Andrew B and Lienig, Jens and Markov, Igor L and Hu, Jin , volume=. 2011 , publisher=

  14. [22]

    2015 , publisher=

    Lu, Jingwei and Zhuang, Hao and Chen, Pengwen and Chang, Hongliang and Chang, Chin-Chih and Wong, Yiu-Chung and Sha, Lu and Huang, Dennis and Luo, Yufeng and Teng, Chin-Chi and others , journal=. 2015 , publisher=

  15. [23]

    2010 , publisher=

    Alagoz, Oguzhan and Hsu, Heather and Schaefer, Andrew J and Roberts, Mark S , journal=. 2010 , publisher=

  16. [24]

    1987 , publisher=

    Cohoon, James P and Paris, William D , journal=. 1987 , publisher=

  17. [25]

    2010 , publisher=

    Van Leeuwen, JMJ , journal=. 2010 , publisher=

  18. [26]

    Chen, Tung-Chieh and Jiang, Zhe-Wei and Hsu, Tien-Chang and Chen, Hsin-Chen and Chang, Yao-Wen , booktitle=

  19. [27]

    Mo, Shibing and Wu, Kai and Gao, Qixuan and Teng, Xiangyi and Liu, Jing , booktitle=

  20. [28]

    Liu, Fei and Tong, Xialiang and Yuan, Mingxuan and Lin, Xi and Luo, Fu and Wang, Zhenkun and Lu, Zhichao and Zhang, Qingfu , booktitle=

  21. [29]

    Achiam, Josh and Adler, Steven and Agarwal, Sandhini and Ahmad, Lama and Akkaya, Ilge and Aleman, Florencia Leoni and Almeida, Diogo and Altenschmidt, Janko and Altman, Sam and Anadkat, Shyamal and others , journal=

  22. [30]

    Liu, Aixin and Feng, Bei and Xue, Bing and Wang, Bingxuan and Wu, Bochao and Lu, Chengda and Zhao, Chenggang and Deng, Chengqi and Zhang, Chenyu and Ruan, Chong and others , journal=

  23. [31]

    Natural Language Analytics with Generative Large-Language Models: A Practical Approach with Ollama and Open-Source LLMs , pages=

    Marcondes, Francisco S and Gala, Adelino and Magalh. Natural Language Analytics with Generative Large-Language Models: A Practical Approach with Ollama and Open-Source LLMs , pages=. 2025 , publisher=

  24. [32]

    Geng, Zijie and Wang, Jie and Liu, Ziyan and Xu, Siyuan and Tang, Zhentao and Kai, Shixiong and Yuan, Mingxuan and Hao, Jianye and Wu, Feng , booktitle=

  25. [33]

    Shi, Yunqi and Lin, Xi and Wang, Zhiang and Xu, Siyuan and Kai, Shixiong and Lai, Yao and Gao, Chengrui and Xue, Ke and Yuan, Mingxuan and Qian, Chao and others , journal=

  26. [34]

    Shi, Yunqi and Lin, Xi and Xu, Siyuan and Kai, Shixiong and Xue, Ke and Yuan, Mingxuan and Qian, Chao and Zhou, Zhi-Hua , booktitle=

  27. [35]

    Cheng, Ruoyu and Yan, Junchi , journal=

  28. [36]

    Xue, Ke and Chen, Ruo-Tong and Lin, Xi and Shi, Yunqi and Kai, Shixiong and Xu, Siyuan and Qian, Chao , journal=

  29. [37]

    Cheng, Ruoyu and Lyu, Xianglong and Li, Yang and Ye, Junjie and Hao, Jianye and Yan, Junchi , journal=

  30. [38]

    2008 , publisher=

    Chen, Tung-Chieh and Jiang, Zhe-Wei and Hsu, Tien-Chang and Chen, Hsin-Chen and Chang, Yao-Wen , journal=. 2008 , publisher=

  31. [39]

    2018 , publisher=

    Cheng, Chung-Kuan and Kahng, Andrew B and Kang, Ilgweon and Wang, Lutong , journal=. 2018 , publisher=

  32. [40]

    2002 , publisher=

    Murata, Hiroshi and Fujiyoshi, Kunihiro and Nakatake, Shigetoshi and Kajitani, Yoji , journal=. 2002 , publisher=

  33. [41]

    Nam, Gi-Joon and Alpert, Charles J and Villarrubia, Paul and Winter, Bruce and Yildiz, Mehmet , booktitle=

  34. [42]

    Xue, Ke and Chen, Ruo-Tong and Tan, Rong-Xi and Lin, Xi and Shi, Yunqi and Xu, Siyuan and Yuan, Mingxuan and Qian, Chao , journal=

  35. [43]

    2008 , publisher=

    Liu, Jing and Zhong, Weicai and Jiao, Licheng and Li, Xue , journal=. 2008 , publisher=

  36. [44]

    L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al

    Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F. L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al. GPT-4 technical report . arXiv preprint arXiv:2303.08774, 2023

  37. [45]

    J., and Roberts, M

    Alagoz, O., Hsu, H., Schaefer, A. J., and Roberts, M. S. Markov decision processes: a tool for sequential decision making under uncertainty . Medical Decision Making, 30 0 (4): 0 474--483, 2010

  38. [46]

    A high-quality mixed-size analytical placer considering preplaced blocks and density constraints

    Chen, T.-C., Jiang, Z.-W., Hsu, T.-C., Chen, H.-C., and Chang, Y.-W. A high-quality mixed-size analytical placer considering preplaced blocks and density constraints . In Proceedings of the 2006 IEEE/ACM International Conference on Computer-Aided Design, pp.\ 187--192, 2006

  39. [47]

    NTUPlace3: An analytical placer for large-scale mixed-size designs with preplaced blocks and density constraints

    Chen, T.-C., Jiang, Z.-W., Hsu, T.-C., Chen, H.-C., and Chang, Y.-W. NTUPlace3: An analytical placer for large-scale mixed-size designs with preplaced blocks and density constraints . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 27 0 (7): 0 12...

  40. [48]

    B., Kang, I., and Wang, L

    Cheng, C.-K., Kahng, A. B., Kang, I., and Wang, L. RePlace: Advancing solution quality and routability validation in global placement . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 38 0 (9): 0 1717--1730, 2018

  41. [49]

    and Yan, J

    Cheng, R. and Yan, J. On joint learning for solving placement and routing in chip design . Advances in Neural Information Processing Systems, 34: 0 16508--16519, 2021

  42. [50]

    The policy-gradient placement and generative routing neural networks for chip design

    Cheng, R., Lyu, X., Li, Y., Ye, J., Hao, J., and Yan, J. The policy-gradient placement and generative routing neural networks for chip design . Advances in Neural Information Processing Systems, 35: 0 26350--26362, 2022

  43. [51]

    Cohoon, J. P. and Paris, W. D. Genetic placement . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 6 0 (6): 0 956--964, 1987

  44. [52]

    EGPlace: An efficient macro placement method via evolutionary search with greedy repositioning guided mutation

    Deng, J., Li, Z., Zhang, J., and Gao, J. EGPlace: An efficient macro placement method via evolutionary search with greedy repositioning guided mutation . In Proceedings of the International Conference on Machine Learning, pp.\ 13237--13255, 2025

  45. [53]

    Reinforcement learning within tree search for fast macro placement

    Geng, Z., Wang, J., Liu, Z., Xu, S., Tang, Z., Yuan, M., Hao, J., Zhang, Y., and Wu, F. Reinforcement learning within tree search for fast macro placement . In Proceedings of the International Conference on Machine Learning, 2024

  46. [54]

    LaMPlace: Learning to optimize cross-stage metrics in macro placement

    Geng, Z., Wang, J., Liu, Z., Xu, S., Tang, Z., Kai, S., Yuan, M., Hao, J., and Wu, F. LaMPlace: Learning to optimize cross-stage metrics in macro placement . In The Thirteenth International Conference on Learning Representations, 2025

  47. [55]

    B., Lienig, J., Markov, I

    Kahng, A. B., Lienig, J., Markov, I. L., and Hu, J. VLSI Physical Design: from Graph Partitioning to Timing Closure , volume 312. Springer, 2011

  48. [56]

    and Markov, I

    Kim, M.-C. and Markov, I. L. ComPLx: A competitive primal-dual lagrange optimization for global placement . In Proceedings of the 49th Annual Design Automation Conference, pp.\ 747--752, 2012

  49. [57]

    MaskPlace: Fast chip placement via reinforced visual representation learning

    Lai, Y., Mu, Y., and Luo, P. MaskPlace: Fast chip placement via reinforced visual representation learning . Advances in Neural Information Processing Systems, 35: 0 24019--24030, 2022

  50. [58]

    Chipformer: Transferable chip placement via offline decision transformer

    Lai, Y., Liu, J., Tang, Z., Wang, B., Hao, J., and Luo, P. Chipformer: Transferable chip placement via offline decision transformer . In Proceedings of the International Conference on Machine Learning, pp.\ 18346--18364, 2023

  51. [59]

    Lin, Y., Dhar, S., Li, W., Ren, H., Khailany, B., and Pan, D. Z. DreamPlace: Deep learning toolkit-enabled GPU acceleration for modern vlsi placement . In Proceedings of the 56th Annual Design Automation Conference 2019, pp.\ 1--6, 2019

  52. [60]

    Deepseek-v3 technical report

    Liu, A., Feng, B., Xue, B., Wang, B., Wu, B., Lu, C., Zhao, C., Deng, C., Zhang, C., Ruan, C., et al. Deepseek-v3 technical report . arXiv preprint arXiv:2412.19437, 2024 a

  53. [61]

    Evolution of heuristics: towards efficient automatic algorithm design using large language model

    Liu, F., Tong, X., Yuan, M., Lin, X., Luo, F., Wang, Z., Lu, Z., and Zhang, Q. Evolution of heuristics: towards efficient automatic algorithm design using large language model . In Proceedings of the International Conference on Machine Learning, pp.\ 32201--32223, 2024 b

  54. [62]

    Moving block sequence and organizational evolutionary algorithm for general floorplanning with arbitrarily shaped rectilinear blocks

    Liu, J., Zhong, W., Jiao, L., and Li, X. Moving block sequence and organizational evolutionary algorithm for general floorplanning with arbitrarily shaped rectilinear blocks . IEEE Transactions on Evolutionary Computation, 12 0 (5): 0 630--646, 2008

  55. [63]

    ePlace-MS: Electrostatics-based placement for mixed-size circuits

    Lu, J., Zhuang, H., Chen, P., Chang, H., Chang, C.-C., Wong, Y.-C., Sha, L., Huang, D., Luo, Y., Teng, C.-C., et al. ePlace-MS: Electrostatics-based placement for mixed-size circuits . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 34 0 (5): 0 6...

  56. [64]

    MacMillen, D., Camposano, R., Hill, D., and Williams, T. W. An industrial view of electronic design automation . IEEE Transactions on Computer-aided Design of Integrated Circuits and Systems, 19 0 (12): 0 1428--1448, 2000

  57. [65]

    AI alone isn't ready for chip design: A combination of classical search and machine learning may be the way forward

    Majumdar, S., Mallappa, U., and Mostafa, H. AI alone isn't ready for chip design: A combination of classical search and machine learning may be the way forward . IEEE Spectrum, 61 0 (12): 0 38--43, 2024

  58. [66]

    S., Gala, A., Magalh \ a es, R., Perez de Britto, F., Dur \ a es, D., and Novais, P

    Marcondes, F. S., Gala, A., Magalh \ a es, R., Perez de Britto, F., Dur \ a es, D., and Novais, P. Using ollama . In Natural Language Analytics with Generative Large-Language Models: A Practical Approach with Ollama and Open-Source LLMs, pp.\ 23--35. Springer, 2025

  59. [67]

    W., Songhori, E., Wang, S., Lee, Y.-J., Johnson, E., Pathak, O., Nova, A., et al

    Mirhoseini, A., Goldie, A., Yazgan, M., Jiang, J. W., Songhori, E., Wang, S., Lee, Y.-J., Johnson, E., Pathak, O., Nova, A., et al. A graph placement methodology for fast chip design . Nature, 594 0 (7862): 0 207--212, 2021

  60. [68]

    AutoSGNN: automatic propagation mechanism discovery for spectral graph neural networks

    Mo, S., Wu, K., Gao, Q., Teng, X., and Liu, J. AutoSGNN: automatic propagation mechanism discovery for spectral graph neural networks . In Proceedings of the AAAI Conference on Artificial Intelligence, volume 39, pp.\ 19493--19502, 2025

  61. [69]

    VLSI module placement based on rectangle-packing by the sequence-pair

    Murata, H., Fujiyoshi, K., Nakatake, S., and Kajitani, Y. VLSI module placement based on rectangle-packing by the sequence-pair . IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 15 0 (12): 0 1518--1524, 2002

  62. [70]

    J., Villarrubia, P., Winter, B., and Yildiz, M

    Nam, G.-J., Alpert, C. J., Villarrubia, P., Winter, B., and Yildiz, M. The ISPD2005 placement contest and benchmark suite . In Proceedings of the 2005 International Symposium on Physical Design, pp.\ 216--220, 2005

  63. [71]

    and Bagnell, D

    Ross, S. and Bagnell, D. Efficient reductions for imitation learning . In Proceedings of the thirteenth International Conference on Artificial Intelligence and Statistics, pp.\ 661--668, 2010

  64. [72]

    Macro placement by wire-mask-guided black-box optimization

    Shi, Y., Xue, K., Lei, S., and Qian, C. Macro placement by wire-mask-guided black-box optimization . Advances in Neural Information Processing Systems, 36: 0 6825--6843, 2023

  65. [73]

    Re ^2 MaP: Macro placement by recursively prototyping and packing tree-based relocating

    Shi, Y., Lin, X., Wang, Z., Xu, S., Kai, S., Lai, Y., Gao, C., Xue, K., Yuan, M., Qian, C., et al. Re ^2 MaP: Macro placement by recursively prototyping and packing tree-based relocating . arXiv preprint arXiv:2511.08054, 2025 a

  66. [74]

    ReMaP: Macro placement by recursively prototyping and periphery-guided relocating

    Shi, Y., Lin, X., Xu, S., Kai, S., Xue, K., Yuan, M., Qian, C., and Zhou, Z.-H. ReMaP: Macro placement by recursively prototyping and periphery-guided relocating . In 2025 62nd ACM/IEEE Design Automation Conference (DAC), pp.\ 1--7, 2025 b

  67. [75]

    The domino effect

    Van Leeuwen, J. The domino effect . American Journal of Physics, 78 0 (7): 0 721--727, 2010

  68. [76]

    Wang, L.-T., Chang, Y.-W., and Cheng, K.-T. T. Electronic design automation: synthesis, verification, and test . Morgan Kaufmann, 2009

  69. [77]

    Reinforcement learning policy as macro regulator rather than macro placer

    Xue, K., Chen, R.-T., Lin, X., Shi, Y., Kai, S., Xu, S., and Qian, C. Reinforcement learning policy as macro regulator rather than macro placer . Advances in Neural Information Processing Systems, 37: 0 140565--140588, 2024

  70. [78]

    BBOPlace-Bench: Benchmarking black-box optimization for chip placement

    Xue, K., Chen, R.-T., Tan, R.-X., Lin, X., Shi, Y., Xu, S., Yuan, M., and Qian, C. BBOPlace-Bench: Benchmarking black-box optimization for chip placement . arXiv preprint arXiv:2510.23472, 2025

Pith tools

Reviewed June 27, 2026 · model on record in the stance chip above.